[e-hentaiにスクレイピング中にでた警告メッセージ - 座敷牢日誌](http://zashikiro.hateblo.jp/entry/2012/09/27/173402)

e-hentai.coffee

#!/usr/bin/env coffee
# vim: set ft=coffee:

cheerio = require 'cheerio'
http    = require 'http'
url     = process.argv[2]
prefix  = process.argv[3] or './'

unless url
  console.warn '''usage: ehentai url [prefix]
  ex) ehentai http://g.e-hentai.org/g/614160/1303da8542/'''
  process.exit 1

id = url.split('/')[4]

## 画像の固有ページのリンクを集めるめう

dfs = (root) ->
  q = [root]
  ret = []
  keyword = "/#{id}-"
  while q.length > 0
    dom = q.shift()
    if (dom.type is 'tag') and (dom.name is 'a') and dom.attribs?.href?
      href = dom.attribs.href
      ret .push href if href.indexOf(keyword) > 0
    if dom.children?
      q.push c for c in dom.children
  ret

iter = (page, ac, cont) ->
  console.warn '@' + page
  page_url = url + (if page is 0 then '' else "?p=#{page}")
  http.get page_url, (req) ->
    data = ''
    req.on 'data', (frag) -> data += frag
    req.on 'end', ->
      $ = cheerio.load data
      n = ac.length
      addition = dfs $._root
      for a in addition
        ac .push a if not (a in ac)
      if n < ac.length
        iter (page + 1), ac, cont
      else
        cont ac

## 画像の固有ページから画像のurlを見つけるめう
image_url = (img_page, ret) ->
  http.get img_page, (req) ->
    data = ''
    req.on 'data', (frag) -> data += frag
    req.on 'end', ->
      $ = cheerio.load data
      src = $('#img').attr('src')
      ret src

## ここがエントリーポイントめう
do ->
  iter 0, [], (pages) ->
    n = pages.length
    iter2 = (idx) ->
      return if idx >= n
      image_url pages[idx], (url) ->
        console.log url
        iter2 idx + 1
    iter2 0